During the GHAC Arbitration Week 2026, Senior Advocate and King’s Counsel Nakul Dewan highlighted the potential pitfalls for legal practitioners who overly depend on artificial intelligence (AI) at the expense of their own analytical capabilities. Dewan expressed concern that lawyers who allow AI to replace critical thinking and thorough case preparation may falter in high-pressure situations, such as unexpected questions during cross-examinations. He referred to this over-dependence as ‘cognitive offloading.’
Dewan emphasized the necessity for lawyers to have a deep understanding of their case files. He stated, “When you use AI, if you substitute it for your own brainwork, then you have a problem… You should certainly use AI, but you use AI to harness your own critical thinking.”
This discussion was part of a panel on ‘Technology & AI in the Proceedings of International and Domestic Arbitration,’ moderated by Senior Advocate Devang Nanavati. The panel also included Barrister Baiju Vasani from Twenty Essex, ICC Indian Chair Sameer Jain, Barrister Rishabh Gupta, and Aditya Singh, Partner at White & Case.
The consensus among the panelists was that AI tools should support, not replace, human intellect. Gupta noted that AI excels in enhancing efficiency, particularly in processing large volumes of documents. “I personally look at it always as a productivity tool, in much the same way as Microsoft Word would have been a productivity tool 20 years ago,” he remarked.
Discussions also centered on the importance of maintaining transparency and procedural integrity in arbitration. Singh stressed the need for clear boundaries regarding AI usage by counsel and tribunals. He stated, “Disclosure, in my view, becomes imperative when AI is being used by the arbitrators.”
More Read
Vasani added that parties select arbitrators for their specific expertise and insights, not merely for their ability to operate AI tools. “They understand you will do that with AI, not simply by AI. They don’t want you to simply delegate your function away,” he said, reinforcing the idea that technology should complement personal expertise.
Jain cautioned against treating AI as a primary legal source, pointing out its limitations in providing definitive answers. “It’s very difficult for AI to say ‘No.’ It’s very difficult for AI to say that, ‘I do not know this,'” he explained.
